Posterior reversible encephalopathy syndrome and reversible cerebral vasoconstriction syndrome in patients with COVID-19 infection: is there a link? A systematic review and case report analysis

Abstract

During the SARS-CoV2 pandemic, several cases of Posterior Reversible Encephalopathy Syndrome (PRES) and of Reversible Cerebral Vasoconstriction Syndrome (RCVS) in COVID-19 patients have been reported, but the link between these syndromes and COVID-19 is unclear. We performed a systematic review, according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement to evaluate whether SARS-CoV2 infection or the drugs used to treat it could be deemed potential risk factors for PRES or RCVS. We performed a literature search. We found 70 articles (60 on PRES and 10 on RCVS) concerning n = 105 patients (n = 85 with PRES, n = 20 with RCVS). We analyzed the clinical characteristics of the two populations separately, then performed an inferential analysis to search for other independent risk factors. We found fewer than usual PRES-related (43.9%) and RCVS-related (45%) risk factors in patients with COVID-19. Such a low incidence of risk factors for PRES and RCVS might suggest the involvement of COVID-19 as an additional risk factor for both diseases due to its capability to cause endothelial dysfunction. We discuss the putative mechanisms of endothelial damage by SARS-CoV2 and antiviral drugs which may underlie the development of PRES and RCVS.

Introduction

COVID-19 is a disease caused by Severe Acute Respiratory Syndrome Corona Virus 2 (SARS-CoV2), a single-stranded RNA beta coronavirus. Most patients develop mild to moderate respiratory symptoms, while others require medical assistance and hospitalization due to respiratory tract involvement. Several long-term complications have been reported including diabetes mellitus, kidney injury, myocarditis, heart failure, and venous thromboembolism [1, 2]. Neurological complications have been described as well: myalgia, dizziness, headache, anosmia, ageusia, encephalitis, stroke, seizures, and myelitis [3–5] but also Posterior Reversible Encephalopathy Syndrome (PRES) and Reversible Cerebral Vasoconstriction Syndrome (RCVS) [6–8].

PRES and RCVS [9–11] are relatively new clinical-radiological syndromes. More and more cases of these syndromes have been described so far, mainly due to increased awareness of their clinical signs and the widespread use of MRI that allows catching their signs at an early stage. Although their etiologies remain unclear, several studies pointed out the role of intracranial vessels and endothelium in their development [12].

PRES, first described by Hinchey et al. in 1996 [13], causes encephalopathy, disorders of consciousness, seizures, visual disturbances, and headache [14]. It is more common in females and the most frequently recognized risk factors are: hypertension, pre-eclampsia, renal disease, immunosuppressive state, infections, steroids, dialysis, and blood transfusions [15, 16]. The pathophysiology of PRES remains unclear, but it seems to be related to a disorder of cerebral vascular autoregulation [17]. The main proposed theories are: (1) increased blood pressure such as to overcome the auto-regulatory mechanisms of cerebral vascularization and to break the blood–brain barrier (BBB) thus leading to vasogenic edema; (2) disrupted cerebral auto-regulation leading to focal vasoconstriction and hypoxia resulting in cytotoxic edema and cerebral infarction; (3) endothelial dysfunction due to endogenous and exogenous toxins which causes capillary leakage, BBB damage and vasogenic edema [16]. Regardless of its causes, typical neuroradiologic findings are bilateral areas of cerebral white matter edema, mainly in the posterior regions, although in some cases cerebral infarction, intraparenchymal hemorrhage, and subarachnoid hemorrhage may be observed in other sites [17, 18].

RCVS is another clinical-radiological syndrome with uncertain etiology. It has a female predominance and the most frequent neurological symptom is thunderclap headache [19], but other manifestations may occur such as brain edema, stroke, and hemorrhage. Potential triggers of RCVS are, among others, medications and drugs, pregnancy, sexual intercourse, and tumors [20]. Key elements are focal vasoconstrictions of brain vessels, which appear in angiographic studies as intermittent focal narrowing, causing the “sausage on a string appearance”. A proposed model suggests that the disease may be caused by an interplay of endothelial dysfunction, oxidative stress, and sympathetic overactivity which results in BBB disruption and cerebral vascular tone dysregulation [21].

Owing to similar pathophysiological mechanisms (namely the breakdown of the BBB, impaired brain vessel autoregulation, and endothelial dysfunction [22, 23]) and because of their coexistence in some patients [8], some authors suggested that PRES and RCVS are a spectrum of the same pathology [7, 8]. Indeed, the release of factors associated with endothelial dysfunction such as endothelin-1, serotonin, cytokines, vascular endothelial growth factor (VEGF) and the role played by oxidative stress, genetic polymorphisms, and autonomic dysregulation [16, 24] have been observed in both syndromes.

Since the beginning of the COVID-19 pandemic, scattered cases of COVID-19 patients who developed PRES or RCVS [9–11] have been reported, but it was unclear whether they were linked to COVID-19 or they were an epiphenomenon of COVID-19.

The aim of this review was to study the characteristics of the described patients with COVID-19 who developed PRES or RCVS and to assess whether SARS-CoV2 infection or COVID-19 therapies could be deemed potential risk factors for PRES or RCVS. In addition, we examined any other risk factor or associated condition that may be a predisposing condition for these syndromes, independently or together with SARS-CoV2 infection.

Methods

Protocol

The review was performed in accordance with recommendations of the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement [25].

Search strategy

We performed a literature search in Medline (PubMed), Scopus, and Web of Science. The search was limited to papers published between December 31, 2019 (the date of the first notification of COVID-19 to WHO [26]) and November 1, 2022. In the Scopus and Web of Science databases, we restricted the search to articles in the field of medicine. The search terms are listed in Table 1.

Table 1.

Search string for PubMed, Web of Science and Scopus

Domain	Search strings
PRES	(“PRES” OR “Posterior reversible encephalopathy syndrome”)
RCVS	OR (“RCVS” OR “Reversible cerebral vasoconstriction syndrome”)
COVID-19	AND (“COVID-19” OR “Coronavirus” OR “SARS CoV-2”)

Eligibility criteria

The inclusion criteria were (1) patients who developed PRES or RCVS after COVID-19 infection and/or COVID-19 vaccine; (2) case reports, case series (full papers or meeting abstracts), and narrative or systematic reviews only if they included case reports. Exclusion criteria were (1) articles with no history of COVID-19 and/or PRES /RCVS infection, (2) patients with PRES or RCVS prior to COVID-19 infection, (3) articles containing no case reports (reviews without personal case reports, letters, or commentaries).

Study selection

The literature search was performed by 3 independent reviewers (AB, GI, SR). Duplicates were removed. Titles and abstracts were first reviewed and those articles that did not meet the selection criteria were removed. The full texts of included articles were then searched, and if unavailable, the articles were removed. The authors reached a consensus and discrepancies were discussed with a fourth author (FP). The study selection strategy is shown in Fig. 1.

Fig. 1.

Flow diagram of literature search and articles selection, n: number of articles

Data extraction

Patient data were extracted by 3 authors (AB, GI, SR). The extracted data were entered into a database with the following information: type of pathology (PRES or RCVS), age, sex, time elapsed (days) between COVID-19 onset or vaccination and the onset of PRES or RCVS, COVID-19 vaccination, comorbidities, risk factors for PRES or RCVS, systolic blood pressure, mean blood pressure, ventilation, COVID-19 complications, PRES or RCVS symptoms, laboratory tests, imaging findings, COVID-19 therapy, PRES or RCVS therapy, clinical outcome. The qualitative data were then converted into dichotomous variables and entered in two separate databases for statistical analysis, one for PRES and one for RCVS.

Quality of the studies

Given the high likelihood of bias in the case reports, the quality of each study was assessed using a modified score of the standardized tool of Murad et al. [27]. Each article received a score between 0 and 8, with 0 indicating a low study quality and a high risk of bias whereas a score of 8 was assigned to high quality studies with a low risk of bias. Because question number 5 (“Was there a challenge/re-challenge phenomenon?”) could not be answered as it referred to adverse drug reactions, an adjusted tool was used with the highest possible score of 7. Question number 6 (“Was there a dose–response effect”?) was assessed making an accurate description of the time lag between COVID -19 positivity and the onset of PRES or RCVS symptoms. In addition to a quantitative assessment, a qualitative assessment of the risk of bias was made by defining it as low, medium–low, medium, medium–high, and high risk of bias based on the score questions.

Data analysis

Continuous variables were expressed as mean and standard deviation. Categorical variables as frequencies and percentages. Because data on the specific items could not be found in all records, the number of patients in whom data on the above-mentioned variables were found was also specified. Differences between quantitative and qualitative variables were performed with the Mann–Whitney U-test, and associations between qualitative variables were performed with the chi-square test and Fisher's exact test. The analysis was performed using the open-source statistical software JASP (JASP Team, 2022, Version 0.16.3 for Windows).

Results

N = 70 articles were found (n = 60 concerning PRES and n = 10 concerning RCVS). Data of n = 105 patients (n = 85 patients with PRES and n = 20 patients with RCVS) were extracted. First, we studied the clinical characteristics of the two populations separately, then performed an inferential analysis to search for other independent risk factors. Table 2 summarizes the reviewed studies.

Table 2.

Literature search results

Authors	Patient no	Age	Gender	Days between COVID-19 and PRES/RCVS onset	Comorbidities	Systolic Blood Pressure Peak	PRES/RCVS Symptoms	Imaging features	COVID-19 therapy	PRES/RCVS Therapy	Prognosis
PRES
Lallana et al. [9] 2019	1	49	Female	17	None	120	Focal signs, visual disturbances	Asymmetric, P-O lesions. Vasocostriction	Antivirals, hydroxychloroquine, corticosteroids, monoclonal Abs	Ca channel blockers	Focal neurological symptoms
	2	36	Female	9	None	116	Focal signs, alterations of consciousness, seizures	Asymmetric, hemispheric,posterior lesions Hemorrhage	Antivirals, hydroxychloroquine, corticosteroids, monoclonal Abs	ASMs	Focal neurological symptoms
	3	66	Male	22	None	127	Alterations of consciousness, seizures	P-O lesions	Hydroxychloroquine, corticosteroids	ASMs	Good outcome
	4	53	Male	44	None	123	Alterations of consciousness, seizures	P-O lesions	None	ASMs	Good outcome
	5	55	Female	39	HTN, DM, dyslipidaemia, obesity	110	Seizures	Asymmetric, P-O lesions	Antivirals, hydroxychloroquine	ASMs	Good outcome
	6	70	Male	1	DM	167	Focal signs, seizures	P-O lesions	Corticosteroids	ASMs	Death
	7	66	Female	48	HTN, DM, dyslipidaemia	210	Alterations of consciousness, seizures	P-O lesions	Antivirals, hydroxychloroquine, corticosteroids, monoclonal Abs, plasma	ASMs	Good outcome
	8	68	Male	70	N/A	218	Alterations of consciousness, seizures	Asymmetric, hemispheric lesions	Corticosteroids	ASMs	Death
Colombo et al. 2019 [10]	9a	54	Male	> 10	N/A	N/A	Seizures, visual disturbances	Asymmetric, P-O lesions	Antivirals	ASMs	Focal neurological symptoms
	10a	63	Female	> 7	N/A	N/A	Alterations of consciousness, visual disturbances, seizures	P-O lesions	Antivirals, hydroxychloroquine, antithrombotic	ASMs	Good outcome
	11a	64	Female	> 7	N/A	N/A	Focal signs, alterations of consciousness, seizures	Asymmetric, hemispheric, posterior lesions Hemorrhage	Antivirals	ASMs	Good outcome
	12a	64	Male	35	N/A	N/A	Alterations of consciousness, visual disturbances	P-O lesions. Hemorrhage	Antivirals	None	Good outcome
	13a	68	Male	> 10	N/A	N/A	Alterations of consciousness, visual disturbances	Asymmetric, P-O lesions	Antivirals	None	Neurological sequelae
	14	57	Male	9	N/A	N/A	Seizures, visual disturbances	Asymmetric, P-O lesions	None	ASMs	Good outcome
Kishfy et al. 2020 [28]	15	58	Male	26	Dyslipidaemia	N/A	Alterations of consciousness, seizures	Asymmetric, P-O lesions. Hemorrhage	Hydroxychloroquine, monoclonal Abs	ASMs	Good outcome
	16	67	Female	25	HTN, DM	N/A	Alterations of consciousness, seizures	Asymmetric, occipital-cerebellar lesions. Hemorrhage	Antivirals	ASMs	Good outcome
Rogg et al. 2020 [29]	17	59	Male	12	None	173	Alterations of consciousness	Parieto-occipital, capsular, cerebellar lesions	Antivirals	None	Death
Hixon et al. 2021 [30]	18a	69	Female	12	HTN, dyslipidaemia	180- None 84	Alterations of consciousness, visual disturbances, seizures	Asymmetric, P-O lesions	Hydroxychloroquine, corticosteroids, antibiotics	None	Focal neurological symptoms
	19a	55	Female	9	HTN, DM	130–160	Focal signs, visual disturbances, seizures	F-P-O lesions. Hemorrhage	Hydroxychloroquine, corticosteroids, antibiotics	None	N/A
	20a	65	Male	39	HTN, DM	140–90	Seizures	F-P-O lesions	Hydroxychloroquine, corticosteroids, antibiotics	None	Good outcome
Llansò et al. 2020 [31]	21a	66	Female	12	N/A	160	Alterations of consciousness	T-O lesions	Antivirals, hydroxychloroquine, monoclonal Abs, antibiotics	None	Death
Parauda et al. 2020 [32]	22a	64	Male	> 5	N/A	187	Focal signs, alterations of consciosuness, visual disturbances, seizures	Asymmetric, P-O lesions	Hydroxychloroquine, antithrombotic	ASMs	Focal neurological symptoms
	23a	73	Male	42	N/A	212	Focal signs, alterations of consciousness, seizures	P-O lesions	Hydroxychloroquine	ASMs	Good outcome
	24	65	Female	> 7	HTN, DM	190	Alterations of consciousness	P-O lesions	Hydroxychloroquine,antithrombotic	None	Neurological sequelae
	25a	74	Female	22	DM, dyslipidaemia, hypothyroidism	237	Focal signs, alterations of consciousness	Asymmetric, P-O lesions	Hydroxychloroquine, monoclonal Abs	None	Neurological sequelae
Elhassan et al. 2021 [33]	26a	54	Female	31	None	190	Focal signs, alterations of consciosuness, visual disturbances, seizures	P-O lesions	Antibiotics	ASMs	Focal neurological symptoms
	28a	34	Female	3	None	210	Alterations of consciousness, seizures	T-O lesions	Plasma	ASMs	Good outcome
Cisowka et al. 2022 [34]	29	38	Female	3	None	N/A	Focal signs, alterations of consciosuness, visual disturbances, seizures	P-O lesions	None	ASMs	Good outcome
Wijeratne et al. 2021 [35]	30	55	0	7	HTN, DM, dyslipidaemia, obesity, CKD	180	Alterations of consciousness	F-P-O lesions. hemorrhage	None	None	Good outcome
Alnass et al. 2021 [36]	31a	43	0	5	DM	136	Alterations of consciousness, seizures	O lesions	Corticosteroids	None	Good outcome
de Lima et al. 2021 [37]	32a	43	Female	N/A	Sickle cell disease	N/A	Alterations of consciousness, seizures	Hemispheric lesions	None	ASMs	Good outcome
Al Haboob et al. 2021 [38]	33a	11	Female	> 5	Miller Fisher syndrome	138	Focal signs, alterations of consciousness, seizures	P-O lesions	Corticosteroids, antivirals, antibiotics	ASMs	Good outcome
Korkmazer et al. 2021 [39]	34	10	0	N/A	None	135	Seizures	Asymmetric, T-P-O lesions	Corticosteroids, antivirals, antibiotics	ASMs	Good outcome
Kobaidze et al. 2021 [40]	35	90	Female	28	HTN, DM, CKD	227	Seizures	Asymmetric, P-O lesions	None	None	Good outcome
Cariddi et al. 2020	36a	64	Female	25	AF, HTN, dyslipidaemia	150	Focal signs, alterations of consciousness, visual disturbances	F-T-P-O lesions. Hemorrhage	Hydroxychloroquine, antivirals, cobicitstat		Good outcome
Doo et al. 2021 [41]	37a	64	0	N/A	N/A	N/A	Seizures	P-O lesions. Hemorrhage	Hydroxychloroquine, antibiotics	ASMs	Good outcome
Talluri et al. 2021 [42]	38a	70	0	5	HTN, CAD, asthma	160	Alterations of consciousness	O lesions	Monoclonal Abs	None	Death
Yaguchi et al. 2021 [43]	39	90	Female	3	HTN	160	Focal signs, seizures	O lesions	None	ASMs	N/A
Perez et al. 2020 [44]	40a	24	Female	4	N/A	N/A	Focal signs, alterations of consciousness	Asymmetric, P lesions	Antivirals, hydroxychloroquine, monoclonal Abs, antibiotics	None	Neurological sequelae
Sripadma et al. 2020 [45]	41a	25	Female	1	Post-partum	190	Alterations of consciousness, seizures	P-O lesions. Hemorrhage	Antivirals, antibiotics	None	Good outcome
Anand et al. 2020 [46]	42a	61	Female	15	N/A	187	Alterations of consciousness, seizures	P-O lesions. Hemorrhage	Antivirals, anakinra	ASMs	Good outcome
	43a	52	Female	34	CKD, HIV	180	Alterations of consciousness, seizures	Asymmetric, F-P-O lesions. hemorrhage	None	None	Good outcome
Sharma et al. 2022 [47]	44a	52	Female	7	N/A	145	Visual disturbances	F-P-O lesions	Corticosteroids, antivirals	None	Good outcome
Boschetti et al. 2020 [48]	45a	46	Male	22	HTN, DM, obesity	130	Focal signs, alterations of consciousness	F-O lesions	None	None	Focal neurological symptoms
Suwanto et al. 2022 [49]	46	7	Female	N	None	N/A	Visual disturbances	O lesions	None	None	Good outcome
Ioan et al. 2022 [50]	47a	38	Male	10	Liver disease	130	Focal signs, alterations of consciousness	O lesions	Corticosteroids, antivirals	ASMs	Good outcome
Kaya et al. 2020 [51]	48a	38	Male	10	N/A	130	Focal signs	Asymmetric. F-O lesions	Hydroxychloroquine, antivirals, antibiotics, antibiotics	None	Good outcome
Conte et al. 2020 [52]	49a	63	Female	37	HTN	N/A	Focal signs	P-O lesions	Antivirals, antibiotics, anakinra	None	Focal neurological symptoms
Enjuto et al. 2020 [53]	50	74	Male	15	Tumour	150	Focal signs, alterations of consciosuness, visual disturbances, seizures	P-O, cerebellar, brainstem lesions	Antivirals, hydroxychloroquine, corticosteroids	ASMs	Good outcome
	51b	64	Female	10	HTN, migraine	150	Focal signs, visual disturbances, seizures	F-P-O lesions. Vasoconstriction. Hemorrhage	None	ASMs, Ca channel blockers	Death
Hill et al. 2022 [54]	52a	54	Female	31	N/A	190	Alterations of consciousness, visual disturbances, seizures	O lesions	Antibiotics	ASMs	Focal neurological symptoms
Naman et al. 2022 [55]	53a	8	Male	32	Tumour	146	Focal signs, alterations of consciousness	P-O, cerebellar lesions. Hemorrhage	Corticosteroids, antibiotics	None	Death
Nosal et al. 2021 [56]	54	33	Female	8	Puerperium	N/A	Alterations of consciousness, visual disturbances	Asymmetric, F-O-cerebellar lesions	Antibiotics, antithrombotic	None	Good outcome
Lal et al. 2021 [57]	55a	54	Male	53	HTN, DM	180	Focal signs	Asymmetric, periventricular lesions	Hydroxychloroquine, antibiotics	None	Good outcome
Donaldson et al. 2021 [58]	56a	66	Female	40	HTN, DM, dyslipidaemia, obesity	N/A	Alterations of consciousness, visual disturbances	O lesions	Antibiotics	None	Good outcome
Al-Kaisy et al. 2021 [59]	57a	32	Female	16	None	154	Seizures	O lesions	Hydroxychloroquine, antivirals, antithrombotic	ASMs	Good outcome
Djellaoui et al. 2021 [60]	58	69	Female	0	Tumour, CAD	N/A	Seizures	Asymmetric, T-O lesions	None	None	Good outcome
D'amore 2020 [61]	59	64	Female	N/A	HTN	N/A	Visual disturbances	F-T-P-O lesions. hemorrhage	None	None	Good outcome
	60	84	Female	N/A	N/A	N/A	Focal signs	Asymmetric, P-O lesions	None	None	Neurological sequelae
Bal et al. 2021 [62]	61	28	Female	N/A	N/A	130	Seizures, visual disturbances	O lesions	Antivirals	None	Good outcome
Franceschi et al. 2020 [63]	62a	48	Male	18	N/A	180	Alterations of consciousness	Asymmetric, P-O lesions. Hemorrhage	None	None	Neurological sequelae
	63	67	Female	2	HTN, DM, CAD	178	Alterations of consciousness	P-O lesions. Hemorrhage	None	None	Good outcome
Varadan et al. 2021 [64]	64	46	Female	40	Liver disease	N/A	Focal signs, alterations of consciousness	F-P-O-cerebellar lesions. Hemorrhage	Corticosteroids	None	Death
Noro et al. 2020 [65]	65	67	Female	1	HTN	150	Seizures	P-O lesions	None	None	Death
Chavez et al. 2021 [66]	66a	61	Female	26	HTN	140	Alterations of consciousness, seizures	P-O lesions. Hemorrhage	Corticosteroids	None	Good outcome
Suwanto et al. 2022 [49]	67	7	Female	10	None	N/A	Visual disturbances, seizures	Asymmetric, T-P-O lesions	None	None	Focal neurological symptoms
Martins et al. [67]	68a	63	Female	41	HTN	N/A	Seizures	Asymmetric, T-P-O lesions. hemorrhage	antibiotici	None	Good outcome
Shankar et al. 2021 [68]	69a	34	Female	N/A	N/A	210	Seizures	Asymmetric, T-O lesions	Plasma	None	Good outcome
Ghosh et al. 2020 [69]	70	33	Female	0	None	N/A	Visual disturbances	T-P-O lesions	Corticosteroids, antibiotics	None	Good outcome
Arslan et al. 2022 [70]	71	5	Female	7	Tumour	142	Alterations of consciousness, visual disturbances, seizures	T-P-O lesions	Hydroxychloroquine, corticosteroids, monoclonal Abs	None	Good outcome
Gasparotto et al. 2022 [71]	72		Female	3	CKD	180	Alterations of consciousness, seizures	Asymmetric, F-P-O lesions. hemorrhage	None	None	Good outcome
Abdulsalam et al. 2021 [72]	73	46	Male	6	HTN	195	Alterations of consciousness, visual disturbances, seizures	Asymmetric, T-P-O lesions	Antivirals, hydroxychloroquine	None	Good outcome
Kerro 2020 [73]	74	85	Male	2	HTN, DM, CKD	184	Alterations of consciousness	T-P-O lesions	None	None	N/A
Dominguez-Rojas et al. [74]	75a	9	Male	21	N/A	N/A	Focal signs, alterations of consciousness, seizures	T-P-O lesions	Corticosteroids	None	Good outcome
McCullough et al. 2022 [75]	76a	76	Female	1	HTN	192	Alterations of consciousness, seizures	T-P-O lesions	Antivirals	None	Neurological sequelae
Dias et al. 2020 [76]	77			N/A	N/A	N/A	Alterations of consciousness	P-O lesions. hemorrhage	None	None	N/A
	78			N/A	N/A	N/A		Asymmetric, O, cerebellar lesions. hemorrhage	None	None	N/A
Mohsen et al. 2022 [77]	79	30	Female	N/A	SLE	N/A	Seizures	P-O lesions	Antivirals	None	Good outcome
Lui et al. 2021 [78]	80a	49	Male	55	N/A	156	Alterations of consciousness, seizures	Asymmetric, P-O lesions	None	None	Good outcome
Chaudhary et al. 2021 [79]	81	60	Female	14	Guillan Barré syndrome	160	Visual disturbances	Asymmetric, T-O lesions	Igs	None	Neurological sequelae
Wilkerson et al. 2021 [80]	82	52	Female	N/A	N/A	165	Alterations of consciousness, seizures	P-O lesions	None	None	Good outcome
Tarabichi et al. 2021 [81]	83	77	Female	3	N/A	N/A	Seizures	F-P-O lesions	None	None	N/A
Taskin et al. 2022 [82]	84	42	Female	4	N/A	N/A	Alterations of consciousness	No lesions	None	None	Good outcome
Nelabhotla et al. 2022 [83]	85	13	Female	7	N/A	180	Seizures, visual disturbances	P-O lesions	None	None	Good outcome
RCVS
Mansoor et al. 2021 [84]	86	31	Female	0	spina bifida, scoliosis	N/A	Visual disturbances	Asymmetric, F-P-O lesions. vasoconstriction	None	Ca channel blockers	Good outcome
Somdattaa et al. 2021[85]	87	64	Female	N/A	Migraine	150	Focal signs, alterations of consciousness, seizures	F-P-O lesions. Vasoconstriction. hemorrhage	None	Ca channel blockers	Death
Dakay et al. 2020 [86]	88	30	Female	15	Migraine	N/A	Seizures	F-P lesions. Vasoconstriction. hemorrhage	None	None	Good outcome
Srinivasan et al. 2021 [87]	89	18	Male	N/A	N/A	149		No lesions	None	None	N/A
Arandela et al. 2021[88]	90a	62	Female	25	HTN, DM, dyslipidaemia, obesity	145	Alterations of consciousness	F-T-P-O lesions. Vasoconstriction	Corticosteroids	None	Good outcome
	91	39	Female	N/A	HTN, DM, dyslipidaemia, obesity	144	Focal signs	Asymmetric, F lesions. Vasoconstriction. Hemorrhage	Corticosteroids	None	Neurological sequelae
	92a	47	Female	21	HTN, DM, dyslipidaemia, obesity	270	Alterations of consciousness, headache	Asymmetric, F-P lesions. Vasoconstriction	Corticosteroids	None	Death
	93a	55	0	21	None	119	Focal signs	No lesions. Vasoconstriction	Corticosteroids	None	Neurological sequelae
	94	35	0	13	None	149	Headache	No lesions	None	None	N/A
	95	54	Female	2	Dyslipidaemia	148	Headache	Asymmetric, P lesions. Vasoconstriction. Hemorrhage	None	None	N/A
	96a	54	Female	30	HTN; dyslipidaemia	218	Alterations of consciousness, seizures	Asymmetric, F-P-O lesions. Vasoconstriction	None	None	Neurological sequelae
	97	37	Female	21	N/A	140	Headache	Asymmetric, F-P-O lesions. Vasoconstriction. Hemorrhage	Corticosteroids	Ca channel blockers	Death
	98	25	Female	2	HTN, obesity	138	Headache, visual disturbances	F-O lesions. Vasoconstriciton	None	Ca channel blockers	Neurological sequelae
	99a	21	Female	30	HTN	80	Headache	Diffuse lesions. Vasoconstriction	None	None	N/A
Harahsheh et al. 2022 [89]	100	44	Male	N/A	HTN, sarcoma	195		Asymmetric, F-P lesions. vasoconstriction	None	None	Good outcome
Scheer et al. 2021 [90]	101	56	Male	30	N/A	N/A	Headache	Asymmetric, F-P-O lesions. Vasoconstriction. Hemorrhage	None	None	Death
Dutta et al. 2021 [91]	102	38	Male	2	None	124	Headache	No lesions. vasoconstriciton	None	None	Good outcome
	103	74	Female	5	Dyslipidaemia, AF	170	Alterations of consciousness, visual disturbances	O lesions. hemorrhage	Corticosteroids	ASMs	N/A
Sadeghizadeh et al. 2022 [92]	104	10	Female	4	N/A	N/A	Focal signs, visual disturbances, seizures	Multifocal lesions. vasoconstriction	Corticosteroids, monoclonal Abs, Igs	None	Good outcome
	105	6	Female	5	HTN	N/A	None	Multifocal lesions. Vasoconstriciton	Corticosteroids. Igs	None	Good outcome
Finsterer 2021 [11]	106b	38	Female	18	None	N/A	Headache	O lesions. Vasoconstriction	None	None	Focal neurological symptoms

N/A not assessable or not reported; F frontal; P parietal; O occipital; T temporal; Abs antibodies; ASMs anti-seizure medications; Igs immuboglobulins; HTN hypertension; DM diabetes mellitus; AF atrial fibrillation, CAD coronary artery disease

^aIndicates that the patient received mechanical ventilation

^bIndicates that the patient was vaccinsted against COVID-19

Population study

PRES

Demographic data of PRES patients

The mean age was 53.0 years (± 20.8 years), the youngest subject was 5 years old, and the oldest was 94 years old. About 65% of the subjects were women, consistent with the greater predisposition of women to present PRES [13]. The time interval between COVID-19 diagnosis or vaccine administration and the appearance of PRES symptoms could be determined in 66 out of 85 subjects, with a mean of 18.4 days (± 16.3 days), with a minimum of 0 days and a maximum of 70 days. We found 2 articles that describe patients who developed PRES following anti-SARS-CoV2 vaccination [75, 77] (2.4% of patients).

Risk factors

Literature data show that the most common risk factors for PRES are: moderate/severe arterial hypertension and especially blood pressure (BP) fluctuations (reported in 75% of patients with PRES) [93–96], renal failure (30%) [16, 97], septic status [16, 98], autoimmune diseases (Systemic erythematosus lupus-SLE [99, 100] hemolytic uremic syndrome [101], thrombotic thrombocytopenic purpura [102]), eclampsia, oncological diseases and chemotherapy (cyclosporine [103], cisplatin, cyclophosphamide, tetracyclines) [104, 105], immunosuppressive therapy [106, 107], monoclonal antibodies (especially VEGF-inhibitors) [108, 109].

Risk factors for PRES were identified in 56.1% of patients in our sample:

• Blood pressure: A mean peak BP of 165.8 mmHg (± 29.8 mmHg) was reported, with BP values higher than 160 mmHg in 43.5% of the patients (30/69 patients). A history of hypertension was reported in 49 patients (51.0% of this subgroup).

• Renal failure: A history of renal failure or renal failure arising during hospitalization was found in 35.1% (20/57 patients).

• Septic shock and sepsis: The presence of sepsis was described in 14.2% of patients (10/70 patients) and 2.8% of patients (2/70) developed septic shock.

• Drugs: One patient used cyclosporine [70] for oncological disease, another one was on azathioprine and hydroxychloroquine therapy because of SLE [77]. Monoclonals (Tocilizumab and Anakinra) were used in 12 patients (out of 64 assessed, 18.8%), antiviral drugs in 40.6% of patients (26/64 patients) in particular: Lopinavir/Ritonavir in 61.5%, Oseltamivir 7.6%, Darunavir/Cobicistat 3.8%, Favipinavir 7.6%, Remdesivir in 26.9%. Chloroquine was administered to 36.4% of patients (24/66 patients), plasma exchange in 2/69 patients (2.89%), corticosteroids in 19/69 patients (27.5%).

• Malignancy: Four subjects had an oncological history (4.8%) specifically: medulloblastoma [55], hemophagocytic lymphohistiocytosis [70], endometrial carcinoma [60] and multiple myeloma [53].

• Autoimmune diseases: Two patients had SLE (2.4%) [71, 77]. No other autoimmune disease was reported.

Clinical features

Literature data show that the most common symptoms in patients with PRES are altered consciousness (50–80% of patients), seizures (60–75%), headache (50%), visual disturbances (33%), and focal neurological disorders such as paresis and/or coordination disorders in about 10–15% of patients [23]. In our analysis, the clinical features of the disease were as follows: seizures in 63.5% of patients, alterations of consciousness in 62.4%, visual changes (cortical blindness, hemianopsia, palinopsia) in 36.5% of the reported cases, whereas neurological focal signs such as hemiparesis, ataxia or dysmetria and cranial nerve changes occurred in 28.2% of cases.

Neuroimaging

Literature data show the presence of hemorrhage in 10–25% of patients, and segmental vasoconstriction in 30–70% of patients [16]. Lesions are most frequently found in the parieto-occipital areas [16].

In our studied population, both brain MRIs and CTs were performed to diagnose PRES. The localization of the lesions was mainly parieto-occipital (91.7%). Frontal localizations occurred in a minority of cases (20.2%). Radiological features showed symmetry of the lesions in 63.1% and the presence of hemorrhage or hemorrhagic spots in 28.6% of cases. Signs of vasoconstriction were noted in 3.6% of cases (3 patients).

Therapy

There is no specific therapy for PRES, therefore its management is supportive and mainly based on blood pressure lowering and symptomatic treatment. Only a little data on the treatment of hypertension was reported. Anti-seizure drugs were used in 67.1% of patients. In one case, vasodilator therapy with nimodipine was used.

Prognosis

In the literature, the incidence of neurological sequelae is reported to be 10–25%, whereas mortality is 3–6% [16]. As for mortality due to PRES in COVID-19 patients, it was not possible to determine whether either COVID-19 or PRES was responsible for the patients’ death (which occurred in 9 patients, 11.1% of cases). Some of the patients with PRES experienced hemiparesis, alterations in consciousness, hemianopsia or other visual disturbances, and persistent seizures. In general, neurological sequelae were observed in 18 out of 72 patients (25.0%).

Quality of studies

The mean score according to the Murad et al. scale [27] was 5.75, with a minimum of 2 (1 case, 1.2%) and a maximum of 7 (27 cases, 32.1%). The overall risk of bias was assessed as medium–low based on the score questions.

RCVS

Demographic data of RCVS patients

Literature data defines a peak of RCVS onset around 42 years [110]. In our population, the mean age of onset was 38.2 years (± 16.6 years), ranging from 6 to 64 years. 65.0% of the patients were women. Neurological symptoms presented on average 13.8 days after COVID-19 diagnosis (± 11.6 days). We found 2 patients with RCVS post-COVID-19 vaccine (10%) [11, 111].

Risk factors

Blood pressure seems to play a role as it occurs in 11–25% of subjects in literature [110, 112]. The use of vasoactive substances, especially those that cause vasoconstriction (serotonergic antidepressants, migraine medications, drugs), is associated with RCVS in 3–52% of cases [110, 112]. Also, migraine actually seems to be associated with this condition, presenting in 17–27% of subjects [110, 112]. Other associated risk factors are pregnancy [113], eclampsia [114, 115], medications (anti-migraine drugs, erythropoietin, antidepressant cough sedatives, adrenergic agents, chemotherapeutic agents, immunoglobulins, oral contraceptives) [116], neurosurgical procedures, hypercalcemia and porphyria [117], unruptured saccular aneurysms [118], cervical artery dissections [118], cerebral venous thrombosis [119, 120], tumors (carcinoid, carotid paraganglioma, pheochromocytoma) [121]. In our sample, common risk factors for RCVS were found in 45% of subject. In particular, we found:

• Headache

In our studied population, 5 patients out of 14 records (35.7%) had a history of headache.

• Hypertension

A mean blood pressure value of 154.9 mmHg (+/- 46.0 mmHg, range 80-270) was reported. Moderate-severe hypertension occurred in 3 out of 14 patients (21.4%). Hypertension was already reported in the medical history of 7 out of 12 patients (58.3%).

• Other associated conditions

In one patient RCVS presented in the course of carotid dissection [86], in one case it occurred in the course of oncologic pathology (pleomorphic sarcoma) [89]. Other associated diseases were Kawasaki syndrome (1 patient) [92], multisystem inflammatory syndrome in children (MIS-C in 1 patient) [92], and spina bifida (1 patient) [84].

Clinical features

Data from literature show that headache occurs in 90–95% of patients, impaired consciousness in 10–15% of patients, seizures in 0–20%, visual symptoms in 30–40%, and focal neurologic deficits in 9–63% of patients [23]. Data from studied population were consistent with the literature: headache was present in 75% of patients (9/12 patients), impaired consciousness in 11.1% (2/18 patients), focal neurologic symptoms in 38.9% of patients (7/18), seizures in 17.6% of patients (3/17), and visual symptoms in 23.7% of patients (4/17).

Neuroimaging

Literature data show that edema was observed in 15–40% of cases, hemorrhage in 15% of patients, vasoconstriction in 100% of patients and coexisting vascular lesions (dissection or aneurysm) in 20% of patients [23].

In our studied sample, RCVS diagnosis was made by means of MRI, CT, CTA, and angiography. Lesions occurred in the posterior and anterior regions (reported in 80% of the patients) and they were asymmetrical in 53.3% of patients. Hemorrhage (intraparenchymal or subarachnoid) was reported in 37.6% of the patients. Evidence of vasoconstriction, a key element for the diagnosis, was obtained in all patients by means of CTA or angiography.

Therapy

Although in RCVS the administration of vasodilators has not been proved to be effective [122], the use of vasodilator therapy (nimodipine) was described in 3 patients.

Prognosis

Neurologic sequelae are common in RCVS patients: approximately 10–15% of affected individuals had residual deficits [123]. In the studied population, sequelae were reported in 5/13 patients (38.5%), and death occurred in 4/17 patients (23.5%).

Quality of studies

The mean score according to the Murad et al. scale [27] was 4.5, with a minimum of 2 (1 item, 5%) and a maximum of 6 (4 items, 20%). The risk of bias was assessed as “medium” based on the score questions.

Table 3 describes the characteristics of the studied patients affected by PRES, RCVS, and COVID-19.

Table 3.

Reversible cerebral vasoconstriction syndrome (RCVS) and Posterior reversible encephalopathy syndrome (PRES) in COVID-19 patients

RCVS				PRES
Variable	Counts	Total	Proportion	Variable	Counts	Total	Proportion
Demographics				Demographic
Female	13	20	0.650	Female	54	83	0.651
Mean age	38.2 years (± 16.6 years)			Mean Age	53 years (± 20.8 years)
Mean days between COVID-19 and RCVS	13.8 days (± 11.6 days)			Mean days between COVID-19 and PRES	18.4 days (± 16.3 days)
Comorbilities				Comorbilities
Obesity	4	20	0.200	Hypertension	25	49	0.510
Vaccination	2	20	0.100	Diabetes	15	49	0.306
Hypertension	3	14	0.214	Dyslipidemia	10	50	0.200
RCVS risk factors				PRES risk factors
High arterial pressure at onset	7	12	0.583	Renal disease	20	57	0.351
Headache history	5	14	0.357	High arterial pressure at onset	30	69	0.435
Malignacy	1	18	0.056	Malignacy	4	85	0.047
				LES	1	2	0.500
				Sepsis or septic shock	46	82	0.561
COVID-19 severity				COVID-19 severity
Mechanical ventilation	5	12	0.417	Mechanical ventilation	41	52	0.788
RCVS symptoms				PRES symptoms
Headache	9	12	0.750	Consciousness alterations	53	85	0.624
Consciousness alteration	2	18	0.111	Neurological focal symptoms	24	85	0.282
Focal signs	7	18	0.389	Seizures	54	85	0.635
Seizures	3	17	0.176	Visual deficit	31	85	0.365
Visual deficit	4	17	0.235
Neuroimaging features				Neuroimaging features
Occipital	12	15	0.800	Occipital	77	84	0.917
Frontal	12	15	0.800	Frontal	17	84	0.202
Asymmetry	8	15	0.533	Asymmetry	31	84	0.369
Hemorrhage	6	16	0.375	Vasoconstriction	3	84	0.036
				Hemorrhage	24	84	0.286
COVID-19 therapies				COVID-19 therapies
Monoclonal antibodies	1	14	0.071	Antivirals	26	67	0.388
Steroids	7	14	0.500	Monoclonal antibodies	12	67	0.179
Immunoglobulin	2	14	0.143	Chloroquine	24	69	0.348
				Steroids	19	69	0.275
				Plasma exchange	2	69	0.028
RCVS therapy				PRES therapy
Nimodipine	3	10	0.300	Antiepileptics	57	85	0.671
				Nimodipine	1	85	0.012
Prognosis				Prognosis
Neurological sequelae	5	13	0.385	Neurological sequelae	18	72	0.250
Death	4	17	0.235	Death	9	81	0.111
Quality index				Quality index
2	1	20	0.050	2	1	84	0.012
3	1	20	0.050	3	1	84	0.012
4	9	20	0.450	4	9	84	0.107
5	5	20	0.250	5	23	84	0.274
6	4	20	0.200	6	23	84	0.274
				7	27	84	0.321

Inferential analysis

Search for unknown risk factors.

We performed an inferential analysis to evaluate whether other conditions, apart from those that are already known, could be deemed risk factors for PRES and RCVS. Such research stems from the observation that the incidence of the commonly reported risk factors for PRES and RCVS is quite low in our population compared to previously reported data. Indeed, several case reports and clinical studies have shown that patients with PRES (and without COVID-19) have hypertension in up to 75% of cases, renal insufficiency in up to 30% of cases, a history of medication intake in 19% of cases, sepsis or eclampsia in 7%, and autoimmune disease in up to 45% of cases [16]. There is a lack of data on the incidence of risk factors in RCVS, and the percentage values reported in the literature vary greatly between studies [110, 112]. In our analysis of the case reports the commonly reported risk factors were absent in 43.9% of PRES patients (46/82 patients assessed) and in 45.0% of patients with RCVS (9/20 patients).

In this study, we specifically examined the portion of the population that did not have any known risk factor for PRES and RCVS, and we considered the risk factors described in the following sections.

Sub-analysis of the population without risk factors for PRES or RCVS

The population with COVID -19 and PRES without commonly reported risk factors for PRES showed a lower mean age than the population with the risk factors (46.7 years ± 21.8 vs. 56.8 ± 19 years). Sex data were similar in both groups. The number of days between COVID-19 manifestation and the onset of the symptoms of PRES was slightly lower in the population without risk factors (16.5 ± 15.7 days vs. 20.5 ± 16.7 days). Regarding PRES symptoms, visual deficits were more common in the population with risk factors than in the population without them (X² = 6.119, OR = 0.321, 95% CI = 0.112–0.878, p = 0.02). The radiological features of PRES were similar in both groups. There was a significant difference in the number of patients receiving antiviral therapy between the two groups: 26 patients received antiviral therapy (15 patients without risk factors and 11 with risk factors) and 41 did not receive antiviral therapy (11 without risk factors and 30 with risk factors). These data show a significant correlation between antiviral therapy and the absence of risk factors for PRES (X² = 6.382, p < 0.02, OR 0.275, 95% CI: 0.095–0.761) (Fig. 2). Notably, most patients with COVID-19 and PRES who received antiviral therapy did not have any risk factor for this condition (Table 4). These findings will be further investigated in the next section. There was no significant difference in terms of prognosis between the two populations: (PRES risk factors and sequelae: X² = 0.348, p = 0.587; PRES risk factors and death: X² = 0.649, p = 0.488) (Table 4).

Fig. 2.

Antiviral drugs and PRES risk factors mosaic plot (X² = 6.382, p < 0.02, OR 0.275, 95% CI 0.095–0.761). The graph shows how most patients with risk factors for PRES did not use antiviral drugs and how most patients who used antivirals did not have risk factors for PRES. RF: risk factors

Table 4.

Sub-analysis of populations with COVID-19 and PRES, either with or without risk factors for PRES

Variable	Risk Factors	No Risk Factors	Stats
Demographics
Age	56.8 ± 19 years	46.7 years ± 21.8	W = 558.0, p = 0.041
Female	28	23	X2 = 0.38, p= 0.5
Days between COVID-19 and PRES	20.5 ± 16.7 days	16.5 ± 15.7 days	W = 431.5, p = 35
Symptoms
Consciousness alteration	33	19	X2 = 3.13, p = 0.07
Focal symptoms	11	12	X2 = 0.88, p = 0.34
Seizures	31	22	X2 = 0.34, p = 0.55
Visual deficit	12	19	X2 = 6.11, p = 0.02, OR = 0.31, 95% CI: 0.125–0.799
Radiological features
Posterior lesions	42	33	X2 = 0.01, p = 0.95
Anterior lesions	7	9	X2 = 1.23, p = 0.26
Asimmetry	17	13	X2 = 0.01, p = 0.93
Vasoconstriction	2	1	X2 = 0.14, p = 0.70
Haemorrhage	14	10	X2 = 0.06, p = 0.79
COVID-19 therapy
Antivirals	11	15	X2 = 6.38, p = 0.02, OR = 0.26, 95% CI: 0.095–0.761
Monoclonal ab	7	5	X2 = 0.05, p = 0.82
Chloroquine	17	7	X2 = 1.98, p = 0.15
Plasma exchange	3	0	X2 = 0.8, p = 0.037
Sterhoid	9	10	X2 = 2.00, p = 0.15
Prognosis
Sequealae	11	6	X2 = 0.34, p = 0.55
Death	4	5	X2 = 0.649, p = 0.42

Comparison between populations with COVID-19 and PRES either with risk factors or without risk factors for PRES. The two populations show similar clinical and demographic characteristics except for the incidence of visual symptoms and antiviral drugs use. These factors are statistically more represented in the population without risk factors for PRES

Patients with RCVS and COVID -19 without known risk factors for RCVS had a slightly lower mean age than the population with the risk factors (35.3 ± 16 years vs. 41.6 ± 17.7 years). The number of days between COVID -19 infection and the onset of RCVS symptoms was similar in both groups. The proportion of females was 45.4% in the group without risk factors and 88.9% in the group with risk factors. Focal neurological symptoms occurred predominantly in the group with risk factors (6/8 patients versus 1/10). Seizures also occurred more often in patients with risk factors (3/8 vs. 0/9). There were no significant differences in terms of neuroradiological features or COVID-19 therapy. The presence of post-RCVS sequelae and mortality were similar in both groups (Table 5).

Table 5.

Sub-analysis of population with COVID-19 and RCVS either with or without risk factors for RCVS

Variable	Risk factors	No risk factors	Stats
Demographics
Age	35.3 ± 16 years	41.6 ± 17.7 years	W = 37.5, p = 0.38
Female	8	5	X2 = 4.10, p = 0.07
Days between COVID-19 and RCVS	16.1 ± 11.3 days	12.4 ± 12.1 days	W = 23.0, p = 0.47
Symptoms
Consciousness alteration	2	0	X2 = 2.81, p = 0.09
Focal symptoms	6	1	X2 = 7.90, p = 0.005, OR = 27.0, 95% CI:1.979–368.383
Seizures	3	0	X2 = 4.09, p = 0.08
Visual deficit	2	2	X2 = 0.01, p = 0.89
Headache	2	7	X2 = 0.14, p = 0.70
Radiological features
Posterior lesions	7	5	X2 = 0.60, p = 0.43
Anterior lesions	7	5	X2 = 0.60, p = 0.43
Asimmetry	4	4	X2 = 0.07, p = 0.78
Haemorrhage	3	3	X2 = 0.15, p = 0.69
COVID-19 therapy
Monoclonal ab	1	0	X2 = 1.07, p = 0.29
Immunoglobulin	1	1	X2 = 0, p = 1.0
Sterhoid	4	3	X2 = 0.28, p = 0.59
Prognosis
Sequealae	3	2	X2 = 0.12, p = 0.72
Death	2	2	X2 = 0.01, p = 0.89

Comparison between populations with COVID-19 and RCVS either with risk factors or without risk factors for RCVS. The populations show similar clinical and demographic characteristics except for the incidence of visual symptoms, more present in the population with risk factors

Antivirals

A significant association was found between the use of antivirals and risk factors for PRES, indeed patients without risk factors for PRES received more antivirals than patients with PRES risk factors (Fig. 2). Hypertension has been the most common risk factors for PRES. For this reason, we also performed a Mann Whitney U-test to assess the difference in peak blood pressure between patients who used antiviral drugs and patients who did not take antiviral drugs. Blood pressure values in patients on antiviral therapy were lower than in patients not taking antivirals (150.6 ± 28.4 vs. 177.4 ± 29.9, W = 391.5, p = 0.007) (Fig. 3). No significant association has been demonstrated between antivirals and sepsis, septic shock, or renal failure. A similar correlation could not be demonstrated in the group of patients with RCVS, because data on the presence or absence of antiviral treatment was lacking.

Fig. 3.

Raincloud plot of peak systolic pressure (in mmHg) and antiviral drugs. Graph shows the blood pressure values of patients who have undergone antiviral therapy and those who have not. Blood pressure was significantly higher in patients who were not treated with antiviral therapy

We also conducted a correlation study between individual antiviral drugs and risk factors for PRES. Although no significant results were found, it was possible to detect a trend between either viral protease inhibitors Ritonavir (the most used antiviral, 61.5%) or Darunavir intake and the lack of risk factors for PRES (X² = 1.659, p = 0.198—Fig. 4). For antiviral drugs other than protease inhibitors, there also seems to be a trend with no statistical significance (X² = 2.068, p = 0.150).

Fig. 4.

Ritonavir/Darunavir and PRES risk factors mosaic plot (X² = 1.659, OR = 0.485, p = 0.2).The graph shows a trend of correlation between risk factors for PRES and absence of ritonavir or darunavir therapy Legend: Rit/Dar: ritonavir/darunavir, RF: risk factors

Other factors

Correlation studies between other COVID-19 therapies (monoclonal antibodies, chloroquine, plasma exchange, and corticosteroids) and risk factors for PRES did not yield significant results. VEGFR inhibitors have been associated with PRES in the literature. In our studied population, the monoclonal antibodies that were administered were tocilizumab (11 patients) and anakinra (1 patient out of 67 patients) but, because of the small number of patients treated with these drugs, it was not possible to assess whether they acted as independent risk factors for PRES. No other factor was observed that could be considered independent risk factors or cofactors for PRES or RCVS in COVID-19 patients.

We did not find a COVID-19 severity score, but we assumed that the need for mechanical ventilation could be considered as a severity index (78.8% of patients-41/52 patients assessed). Data show no statistically significant correlation or trend between the use of antivirals and mechanical ventilation (X² = 1.681, p = 0.195) suggesting how antiviral drugs use was not associated with COVID-19 severity in the studied population.

Discussion

We analyzed the clinical features of reported patients who developed PRES and RCVS after SARS-CoV-2 infection. These patients had an incidence of symptoms, neuroradiologic features, and long-term outcomes consistent with the diagnosis of PRES and RCVS as previously described in literature [23]. However, COVID-19 patients reported a lower incidence of predisposing risk factors for PRES and RCVS [16, 23]. Such a low incidence of common risk factors for RCVS and PRES in these patients might be related to COVID -19 infection which may behave as an adjunctive or independent risk factor for both diseases.

COVID-19 and PRES or RCVS

This relationship may be supported by the effects of COVID -19 on the vascular endothelium [124], with the impairment of cerebral autoregulatory mechanisms representing the pathophysiological process of both RCVS and PRES [17, 21]. Some studies reported that SARS-CoV2 may enter the CNS [125, 126] and exert its effects on the endothelium [124], while others demonstrated that it binds to the angiotensin-converting enzyme 2 (ACE2) receptor, which is highly expressed in neurons and glial cells [125]. The underlying pathophysiological mechanism of both diseases seems to be related to an altered cerebrovascular regulation and, on this ground, some authors consider them as a spectrum of the same pathology [7, 8]. As a matter of fact, COVID-19 may predispose to both diseases by means of the endothelial dysfunction caused by SARS-CoV2 [124]. The first pathogenic moment is triggered by the viral invasion of mucosal epithelial and endothelial cells. Penetration is facilitated primarily by ACE-2, a key element of the renin–angiotensin–aldosterone system responsible for converting angiotensin II to Angiotensin-1–7, which has anti-inflammatory, endothelial-protective, and vasodilator properties. Some studies have demonstrated an increased expression of ACE-2 and associated endothelitis in autopsies from COVID-19-positive individuals compared to healthy individuals [127], although other studies showed a low expression of ACE-2 [128] in human endothelium, resizing the pathogenic role of this receptor.

The alteration of the endothelium may be the result of several mechanisms:

• Endothelial cell damage and apoptosis [129]: caused by hyperinflammation and activation of the inflammasome by cytokines (IL-6, TNF-alpha, ICAM-1, caspase-1);

• Damage and degradation of endothelial proteoglycan [130]: caused by the binding of the Spike protein to the proteoglycan that compromises the integrity of the extracellular matrix;

• Hyperpermeability of the endothelium: caused by inflammation and disruption of junctional proteins [131] due to cytokine/chemokine storm and adhesion molecules;

• Inflammation and adhesion of leukocytes to the endothelium [132]: due to the S protein, which increases the release of inflammatory molecules and of molecules that stimulate leukocyte adhesion (THP-1);

• Increased endothelial stress and decreased bioavailability of nitric oxide[133]: through ROS -mediated activation of NADPH oxidase leading to the uncoupling of eNOS;

• Virus-induced senescence of the endothelium 134: due to the increased release of VEGF which stimulates angiogenesis (leading to endothelial leakage and inflammation) and through premature aging of endothelial cells;

• Activation of complement [135]: SARS-CoV-2 increases C3a levels, which triggers the activation of CD16 + T cells that promote endothelial damage.

Most of the pathogenetic mechanisms are mediated by the SARS-CoV-2 spike protein (protein S) that is capable of damaging the endothelium by several mechanisms: it stimulates the release of procoagulant substances (tissue factor) and cyto-chemokines (ICAM-1, VCAM-1, MCP-1) [136]; it causes the release of proinflammatory molecules (IL -6, IL -18, MCP-1, PAI-1) [137] and it leads to the degradation of junctional proteins, resulting in endothelitis and fluid extravasation [138, 139].

Antiviral drugs and PRES in COVID-19 patients

We also found a significant correlation between PRES and the use of antiviral drugs in patients with COVID-19, who do not have other known risk factors for PRES.

Several explanations for this correlation may be proposed. Firstly, antivirals may have been used in patients with more severe COVID-19 symptoms, although we did not find a correlation with ventilation. Thus, inflammatory response to COVID-19 would predispose to endothelial dysfunction and thus to PRES; this association could be an epiphenomenon of COVID-19 severity. Secondly, antivirals could lead to biological effects that result in a reduction of risk factors. In this case, the lower incidence of risk factors in the population would be explained by the effect of antiviral on modifiable risk factors: blood pressure, new-onset renal failure, and septic shock. Indeed, the presence of a strong statistical correlation between blood pressure values and antiviral therapy could be because of such drugs. However, this hypothesis does not appear to have a biological substrate. Ritonavir, the most commonly used antiviral in this population, has not been shown to have antihypertensive effects; on the contrary, it has been associated with hypertensive effects [140] and remdesivir and other antivirals have also never shown antihypertensive effects.

Finally, antivirals may act as an independent risk factor or as a cofactor for PRES development. The higher use of antivirals as part of the therapy for COVID-19 in patients without known risk factors for PRES and the fact that some of them then developed PRES, may mean that antivirals, alone or in combination with COVID-19, are themselves a risk factor for PRES. This hypothesis has an important biological substrate. The most used antivirals were ritonavir (61.5%) and remdesivir (26.9%). The former is mainly used to treat HIV infection and its function is to inhibit the viral protease. Several studies have demonstrated the cytotoxic effects of ritonavir on the vascular endothelium [141–143]. The drug seems to reduce the adiposity of the endothelium and leptin secretion, which results in a reduction of endothelial leptin signaling and, consequently, in a decreased bioavailability of NO, which is induced by NADPH oxidase 1 and mediated by CCR5 [142].

The result is an increased vascular adrenergic contractility and endothelial dysfunction [142]. In vivo studies in HIV patients receiving ritonavir therapy and in healthy patients receiving ritonavir therapy show that this dysfunction occurs primarily in individuals with HIV [144]. In addition to this, there are some case reports of HIV-positive patients under treatment with ritonavir who developed PRES even if they did not have any risk factor for it [145]. Darunavir, which share the same mechanism of action of ritonavir, also appears to be associated with endothelial alterations and increased risk of thrombosis [146] and we found a trend between ritonavir or darunavir intake and the absence of PRES risk factors.

Remdesivir is a nucleotide analogue, mainly used against Ebola virus [147], which has been associated with cardiac and renal toxicity [148] and neurotransmitter release disruption [149]. However, there is no evidence of direct endothelial damage in individuals treated with remdesivir. Oseltamivir and favipavir, which act as neuroamidase inhibitors, are drugs used against influenza viruses, but studies did not show any evidence of endothelial damage. We found no case reports about patients developing PRES while taking antivirals except in a study in which ritonavir was used in a patient with HIV [145]. It is conceivable that these drugs act as cofactors together with SARS-CoV-2 rather than as independent risk factors for PRES development.

COVID-19, antivirals and endothelial dysfunction

Our analysis shows an association between COVID -19 patients undergoing antiviral drugs (especially viral protease inhibitors) and PRES and RCVS. This association might have a biological substrate, and the possible pathogenetic mechanism underlying this association may be the endothelial dysfunction caused by COVID-19 and antivirals. Indeed, both induce NADHP oxidase 1 and reduce endothelial NO bioavailability [133, 142]. NO is an endogenous mediator of several physiological processes at the endothelial level and regulates vascular homeostasis: it has vasodilatory, antithrombotic, and anti-inflammatory effects. Alteration of the bioavailability of this molecule leads to vasoconstriction, thrombosis and increased ROS production causing increased oxidative stress and direct damage to the endothelium [150]. Therefore, it is reasonable to assume a convergence and potentiation of the deleterious effects on the endothelium by COVID -19 and protease inhibitors (Fig. 5).

Fig. 5.

Conceivable pathogenic mechanism. Both Covid-19 and protease inhibitors may contribute to endothelial disruption, by reducing endothelial Nitric Oxide (NO) bioavailability. These alterations may lead to the onset of PRES or RCVS

The occurrence of PRES in patients with COVID-19 and the absence of other risk factors for PRES suggests a link between these diseases and this correlation appears to be consistent. This consideration is less evident for RCVS patients, due to the small sample size and limited knowledge about its predisposing risk factors.

To our knowledge, this is the first study to highlight the relationship between COVID-19, antivirals, and PRES.

Significance of these findings and prospects

New COVID-19 variants have been shown to be more diffusive but less aggressive leading to a substantial reduction in complications. Furthermore, they seem to show lower endothelial toxicity [151, 152]. However, the importance of these findings is not diminished, because they may clarify some of the molecular mechanisms implicated in these brain vascular diseases. Another important role seems to be played by antiviral drugs (especially protease inhibitors), as they are used not only in COVID-19 but also for the treatment of other diseases. Their use in patients who already have other risk factors for PRES could be associated with increased damage to the endothelium and thus with an augmented risk of developing the disease. Further studies in these patients are warranted in order to assess if they have a higher risk of developing PRES.

Limitation

We are aware of the limitations of this study. The analysis relies on case reports which do not always provide complete information. The quality of the studies was assessed using a semi-quantitative scale (Murad et al.) [27] which showed a higher risk of bias for case reports on RCVS (mean score 4.5/8, medium risk of bias) and a lower risk for reports about PRES (mean score 5.75/8, medium–low risk of bias). Overall, the case reports regarding PRES were more detailed, containing more anamnestic, clinical, and therapeutic information, allowing a more reliable analysis between PRES and COVID -19 population. Another limitation is the absence of a control group, which results in a low level of evidence for COVID-19 and antivirals as independent risk factors for PRES, even if, considering the relative rarity of these diseases, it is extremely difficult to conduct a randomized trial with control groups.

Conclusion

We found an association between PRES, COVID-19, and antiviral drug therapy for COVID-19. These findings might be due to a direct effect of SARS-COV-2 on intracerebral circulation.

However, these findings should be taken with caution because no definite relationship may be traced before a larger study is conducted. Further studies are warranted with larger samples and randomized control groups to evaluate whether other viruses or antivirals can lead to increased risk for PRES and RCVS even in the absence of COVID-19 infection.

Funding

None.

Data availability

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Declarations

Conflicts of interest

None.